SEE COMMENTARY

RUNX1 is essential for mesenchymal stem cell proliferation and myofibroblast differentiation Woosook Kima, David A. Barronb,1, Rebeca San Martinb, Keith S. Chana,b,c, Linda L. Trana, Feng Yangb, Steven J. Resslerb,2, and David R. Rowleya,b,3 a Integrative Molecular and Biomedical Sciences Graduate Program, bDepartment of Molecular and Cellular Biology, and cScott Department of Urology, Baylor College of Medicine, Houston, TX 77030

Myofibroblasts are a key cell type in wound repair, cardiovascular disease, and fibrosis and in the tumor-promoting microenvironment. The high accumulation of myofibroblasts in reactive stroma is predictive of the rate of cancer progression in many different tumors, yet the cell types of origin and the mechanisms that regulate proliferation and differentiation are unknown. We report here, for the first time to our knowledge, the characterization of normal human prostate-derived mesenchymal stem cells (MSCs) and the TGF-β1–regulated pathways that modulate MSC proliferation and myofibroblast differentiation. Human prostate MSCs combined with prostate cancer cells expressing TGF-β1 resulted in commitment to myofibroblasts. TGF-β1–regulated runt-related transcription factor 1 (RUNX1) was required for cell cycle progression and proliferation of progenitors. RUNX1 also inhibited, yet did not block, differentiation. Knockdown of RUNX1 in prostate or bone marrow-derived MSCs resulted in cell cycle arrest, attenuated proliferation, and constitutive differentiation to myofibroblasts. These data show that RUNX1 is a key transcription factor for MSC proliferation and cell fate commitment in myofibroblast differentiation. This work also shows that the normal human prostate gland contains tissue-derived MSCs that exhibit multilineage differentiation similar to bone marrow-derived MSCs. Targeting RUNX1 pathways may represent a therapeutic approach to affect myofibroblast proliferation and biology in multiple disease states. myofibroblast

|

MSC

|

RUNX1

|

TGF-β1

|

is found in most cancers and is typified by the coevolution of myofibroblasts (5–7). Importantly, the volume of reactive stroma relative to cancer is predictive of the rate of cancer progression in several tumor types (8–15). Furthermore, in vivo modeling has shown that reactive stroma is tumor-promoting (16–19). The myofibroblast is the principal cell type in reactive stroma, and transforming growth factor beta 1 (TGF-β1) is a potent inducer of myofibroblast differentiation (3, 20). TGF-β1 is overexpressed in most human cancers, including prostate cancer (2, 7). However, the cell types of origin and mechanisms that regulate their proliferation and differentiation to myofibroblasts remain unknown. To address tissue-resident MSCs in the prostate gland and their potential induction to myofibroblasts, we have evaluated whether tissue-associated MSCs reside in normal adult human prostate tissue, whether they are induced to become myofibroblasts by TGF-β1, and the key mechanisms that regulate their proliferation. We report here for the first time, to our knowledge, the isolation of cells with prototypical MSC properties derived from normal adult human prostate gland. These cells exhibit multipotent differentiation similar to bone marrow-derived MSCs and form typical reactive stroma myofibroblasts under TGF-β1 regulation. Importantly, we have identified runt-related transcription factor 1 Significance

reactive stroma

Recruitment, proliferation, and differentiation of myofibroblasts are common in many disease states. Mechanisms that regulate proliferation and differentiation are poorly understood, although TGF-β is a key inducer of differentiation. Here, we report, for the first time to our knowledge, that runt-related transcription factor 1 (RUNX1) regulates mesenchymal stem cell (MSC) biology and progenitor cell commitment to myofibroblasts. In this work, we describe the first identification, to our knowledge, of tissue-resident MSCs from adult normal human prostate gland and the role of these MSCs as myofibroblast precursors. We also pinpoint the role of RUNX1 in regulating proliferation and differentiation in both marrow-derived and tissue-resident MSCs. Perturbation of RUNX1 activity may provide insights for developing antifibrotic and anticancer therapies via targeting the reactive stroma microenvironment.

M

yofibroblasts are key mediators of homeostasis-associated biology in reactive stroma associated with tissue repair, fibrosis, and the tumor microenvironment. However, the specific cell types of origin and the mechanisms that regulate their activation, proliferation, and differentiation are not understood. Accordingly, identifying the niche and origins of myofibroblasts and pathways that regulate their proliferation and commitment to myofibroblasts is important to developing antifibrotic and anticancer therapeutic approaches designed to target the evolution and proliferation of myofibroblasts. Adult mesenchymal stem cells (MSCs) are defined as multipotent stromal cells capable of self-renewal and differentiation to cartilage, bone, and adipose tissues (1). Studies have primarily addressed the biology of marrow-derived MSCs and far less is understood about putative tissue-resident MSCs and how they contribute to the biology of their tissue of residence. Recent evidence from mouse models implicates tissue-resident MSCs in repair processes of disrupted local tissue homeostasis, suggesting that they are a critical component in wound repair, fibrosis, cardiovascular disease, and in the tumor microenvironment (2). Each of these is associated with evolution of reactive stroma with expression of repair-centric genes. However, very little is known about adult human tissue-resident MSCs and the mechanisms that regulate their proliferation and lineage commitment to local reactive stroma or fibrotic tissue. Reactive stroma is dynamic and responds rapidly to emerging situations to restore disrupted homeostasis (3, 4). Reactive stroma

www.pnas.org/cgi/doi/10.1073/pnas.1407097111

Author contributions: W.K., K.S.C., and D.R.R. designed research; W.K., D.A.B., R.S.M., L.L.T., and F.Y. performed research; W.K., S.J.R., and D.R.R. analyzed data; and W.K. and D.R.R. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Freely available online through the PNAS open access option. See Commentary on page 16238. 1

Present address: Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY 10065.

2

Present address: Aptalis Pharmaceuticals, Bridgewater, NJ 08807.

3

To whom correspondence should be addressed. Email: [email protected].

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. 1073/pnas.1407097111/-/DCSupplemental.

PNAS | November 18, 2014 | vol. 111 | no. 46 | 16389–16394

CELL BIOLOGY

Edited by Owen N. Witte, Howard Hughes Medical Institute, University of California, Los Angeles, CA, and approved September 22, 2014 (received for review April 18, 2014)

(RUNX1) as a key transcription factor that is required for MSC/ progenitor cell proliferation and that functions to moderate myofibroblast differentiation. Results Prostate-Derived CD44+CD90+ Cells Are Functional Mesenchymal Stem Cells. Fresh tissue cores from the peripheral zone of nor-

mal human prostate glands were obtained from 19-y-old (HPS-19I cells) and 33-y-old (HPS-33Q cells) cadaver donors with no prostate gland histopathology and explants cultured in a selective medium and protocol optimized for the culture of fetal rodent urogenital sinus mesenchyme (21). The resulting monolayers were evaluated for phenotypic markers and biology. Both HPS-19I and HPS-33Q cells exhibited spindle-shaped fibroblast-like morphology (Fig. 1A) and were capable of long-term culture for more than 25 passages, indicating a long-term growth potential. These cells exhibited density-dependent inhibition of cell growth when reaching more than 70% confluency, which restricted subsequent passage efficiency and differentiation potential. Both HPS-19I and HPS-33Q cells exhibited a normal diploid karyotype and lacked chromosomal aberrations, including translocations as indicated by spectral karyotyping (SI Appendix, Fig. S1). Flow cytometry showed that both cell lines were positive for CD44 and CD90 (Fig. 1B). To assess spatial distribution of CD44+CD90+ cells in prostate tissue, a tissue array with 32 normal human prostate tissue cores was evaluated via multispectral immunohistochemistry (IHC). Distinct clusters of CD44+CD90+ dual positive cells and individual cells were localized in the stromal compartment (Fig. 1C). In addition, dual positive immunoreactivity was observed in small vessels of the microvasculature and nerves. Although many fields exhibited a density of 1% or lower, the average density was 2.83% of total cells (SI Appendix, Fig. S2).

HPS-33Q

HPS-19I

HPS-33Q

CD44 :: PE CD44 CD90

Osteocalcin

E

HPS-19I (p9) HPS-33Q (p7) hBM-MSC (p4)

Pseudo

Pseudo

F

CD44

CD73

CD90 CD29

CD13 CD106 CD105 STRO-1

G CD106

CD44

CD29 CD90

CD105 BMP2

CD73

CD13

OCT4 CD133

Log10 hBM-MSC 2-∆Ct

HPS-19I

hBM-MSC CD29 :: PE

CD90 :: FITC

CD90 :: FITC

CD73 :: FITC CD105 :: PE

CD44 CD90

Aggrecan

Neurofilament M

CD44 :: PE

C

D

CD13 :: PE

B

HPS-19I

Log10 HPS-19I 2-∆Ct

A

CD106 :: FITC STRO-1 :: FITC

Fig. 1. CD44+CD90+ stromal cells derived from normal human prostate are mesenchymal stem cells. (A) Cell morphology. (Scale bars, 100 μm.) (B) Flow cytometric analysis of HPS cell lines. (C) Immunostaining of normal human prostate tissue for CD44 (blue), CD90 (red), and nuclei (green). Multispectral deconvolution microscopy revealed localization of CD44+CD90+ cells (“Pseudo” yellow, arrows) in the stroma. (Scale bars, 100 μm.) (D) Multilineage differentiation potential of HPS-19I cells. Neurogenesis is shown by neurofilament M staining (“Neurofilament M”). (Scale bars, 100 μm.) Chondrogenesis by aggrecan accumulation (“Aggrecan”). (Scale bars, 20 μm.) Osteogenesis by osteocalcin production (“Osteocalcin”). (Scale bars, 20 μm.) (E) Flow cytometric analysis of cell lines HPS-19I, HPS-33Q, and hBM-MSCs. “p” denotes passage. The open and gray plots represent the isotype control and the specific antibody indicated, respectively. (F) Similar gene expression profiles of HPS-19I cells compared with hBM-MSCs. (G) Flow cytometric analysis revealed that HPS-19I cells and hBM-MSCs share expression of MSC surface markers.

16390 | www.pnas.org/cgi/doi/10.1073/pnas.1407097111

Owing to the expression of MSC-associated markers, longterm passage potential, and diploid status, we addressed whether these cells exhibit multipotent MSC characteristics. To address multilineage differentiation potential, HPS-19I and HPS-33Q cells were evaluated relative to human bone marrow-derived MSCs (hBM-MSCs). HPS-19I cells were induced to differentiate into neurogenic, chondrogenic, and osteogenic lineages (Fig. 1D), but were restricted from adipogenic differentiation. hBM-MSCs differentiated into all four lineages (SI Appendix, Fig. S3A). HPS-33Q cells could be induced to chondrocytes (SI Appendix, Fig. S3B), whereas osteogenic induction was less clear. Expression profiles of MSC-specific markers and stem cell markers in HPS-19I cells were congruent with profiles in hBM-MSCs (Fig. 1F). Moreover, flow cytometry revealed that the cell lines HPS-19I, HPS-33Q, and hBM-MSCs were each positive for CD44, CD90, CD13, CD29, CD73, and CD105, the most commonly reported human MSC markers, and were low for CD106 and STRO-1 (Fig. 1 E and G). Based on specific morphology, plastic adherence, multipotency, transcriptome, and cell-surface protein profiles of HPS-19I and HPS-33Q cells, we hereafter refer to these cells collectively as human prostate-derived mesenchymal stem cells (hP-MSCs). Carcinoma hP-MSC Cell Interactions in 3D Organoid Cultures. A 3D organoid coculture system in defined medium that permitted the study of cell–cell interactions and differentiation potentials was developed (Fig. 2A). A combination of LNCaP human prostate carcinoma cells with hP-MSCs resulted in free-floating self-organizing organoids exhibiting a central core of hP-MSCs and an outer mantel of LNCaP (Fig. 2B). Organoids constructed with LNCaP and hBMMSCs exhibited an identical phenotype (SI Appendix, Fig. S4A). IHC of 3D organoids showed that the core of hP-MSCs was positive for CD44 and CD90 and that the LNCaP mantel was positive for androgen receptor (AR) (Fig. 2C). LNCaP or hP-MSCs seeded alone also self-organized as a free-floating 3D organoid (SI Appendix, Fig. S4A) and could be indirectly cocultured with other cells as monolayers. Combination of LNCaP and hP-MSCs inoculated as s.c. xenografts in nude mice exhibited a similar self-organization in vivo with typical carcinoma foci and adjacent stroma (Fig. 2B; SI Appendix, Fig. S4B). To investigate the function of TGF-β1 and hP-MSCs in the genesis of reactive stroma, hP-MSCs were cocultured with LNCaP engineered to express constitutively active TGF-β1 (SI Appendix, Fig. S6A). Organoids constructed with TGF-β1–expressing LNCaP exhibited elevated immunoreactivity for α-smooth muscle actin (α-SMA) and tenascin-C in hP-MSCs relative to control (Fig. 2D). Both are prototypical markers of reactive stroma myofibroblasts (7). Tenascin-C was predominantly localized in stroma immediately adjacent to epithelial cells, similar to patterns noted in human prostate disease (7). In addition, hP-MSCs exhibited increased immunoreactivity for both vimentin and fibroblast activation protein (FAP) under TGF-β1–stimulated conditions (SI Appendix, Fig. S5A). Furthermore, in vivo xenografts (day 11) generated with hPMSCs combined with LNCaP resulted in a central core of α-SMA– positive myofibroblasts (SI Appendix, Fig. S4C). Labeling of the hP-MSCs via expression of RFP marker protein showed RFP and α-SMA dual positive cells in combined xenografts (day 10) (SI Appendix, Fig. S13). TGF-β1–Induced Gene Expression Profiles of hP-MSCs Exhibit a Reactive Stroma Myofibroblast Signature. To identify gene ex-

pression profile alterations, we analyzed indirect cocultures of hPMSC and engineered LNCaP organoids (Fig. 2A). ELISA analysis confirmed that active TGF-β1 was observed only in cocultures made with TGF-β1–expressing LNCaP (SI Appendix, Fig. S6B). hP-MSCs cocultured with LNCaP expressing TGF-β1 exhibited an increased expression of reactive stroma genes such as TNC, ACTA2, FAP, and COL1A1 and showed distinct gene profiles including genes encoding cell-surface markers and growth factors Kim et al.

Transcription Factors Associated with MSC Proliferation and Myofibroblast Differentiation. To better understand mechanisms

that may regulate proliferation and myofibroblast differentiation, Kim et al.

A

Relative expression to GAPDH

B 0.008

Merge

RUNX1

DAPI

siNC siRUNX1

0.006 0.004 0.002 0.000

* * * *** 24 h 48 h 72 h 96 h

RUNX1

X1

to r

N U

Ve c

R

C

X1

R si

WNT11

X1 N

to r

U

Ve c

*

R

N

X1

C

***

*

U

si N

**

*

*

R

U

N

X1

r to

N

si N

r

X1

to

N U

Ve c

R

8 6 4 2 0

Ve c

N

Control TGF- 1

*** **

PTGS2 **

X1

C N

** ***

20 15 10 5 0

U

*

R

C

X1

N

N

si

si R **

*

si

X1 N

25 20 15 10 5 0

U

**

ACTA2

***

U

X1 N U

**

**

**

R

Ve ct or

***

40 35 30 15 10 5 0

BMP6 **

**

U

10 8 6 4 2 0

r

-Actin

Vector RUNX1

to

RUNX1

*

**

Ve c

0.00

** ***

R

0.01

C

***

0.02

*

6 4 2 0

X1

0.03

**

**

TNC

**

si

RUNX1

D8

C

-Actin

Merge

RUNX1

si R

DAPI

N

important in stromal-epithelial interactions, including CTGF, IGF1, BMP6, IL8, IL6, WNT5A, and WNT11 (Fig. 2E; SI Appendix, Fig. S7). Several pathway-focused PCR arrays revealed differential regulation of genes involved in a range of biological processes likely to be important in carcinoma–myofibroblast interactions (SI Appendix, Dataset S1). Direct-contact monolayer cocultures and 3D organoids were also analyzed to assess how cell-to-cell contact affects gene expression. Direct cocultures of RFP+ hP-MSCs and GFP+-engineered LNCaP showed that fluorescent-activated cell sorting (FACS)-sorted hP-MSCs exhibited a similar TGF-β1–regulated gene expression profile as in the indirect cocultures (SI Appendix, Fig. S8). Furthermore, laser capture microdissection was used to harvest RNA from hP-MSCs in 3D organoids with TGFβ1–expressing LNCaP. hP-MSCs in 3D organoids exhibited a distinct reactive stroma gene signature (SI Appendix, Fig. S5B), consistent with the protein expression data (Fig. 2D; SI Appendix, Fig. S5A). In addition, PC3 human prostate cancer cells naturally express and secrete latent TGF-β1. Conditioned media from PC3 cells (TGF-β activated by acid) induced a nearly identical gene expression pattern in hP-MSCs, as observed with engineered LNCaP, and this response was abrogated by SD208, a TGF-β type I receptor blocker (SI Appendix, Fig. S9). Together, these results further support the concept that TGF-β1–expressing prostate cancer cells induce hP-MSCs to a myofibroblast differentiation pattern and mediate the reactive stromal response via altered expression of a wide range of growth factors and cytokines.

Fig. 3. Loss of RUNX1 shows distinct gene expression patterns in hP-MSCs (HPS-19I), which is potentiated by TGF-β1. (A) Immunofluorescent staining of RUNX1 in control and TGF-β1–treated hP-MSCs. DAPI was used to visualize cell nuclei. Note nuclear localization of RUNX1 in TGF-β1–stimulated conditions. (Scale bars, 100 μm.) (B) qRT-PCR and Western blot of RUNX1 in hP-MSCs transfected with control or RUNX1 siRNA. (C) qRT-PCR and Western blot of RUNX1 in control and RUNX1-overexpressing hP-MSCs. (D) Gene expression analysis in RUNX1 knockdown or overexpressing hP-MSCs cocultured with LNCaP control or LNCaP-expressing TGF-β1. Note that depletion of RUNX1 exhibited a myofibroblast signature in hP-MSCs, which was augmented by TGFβ1. Data represent mean ± SEM. n = 3; *P < 0.05; **P < 0.01; ***P < 0.001.

PNAS | November 18, 2014 | vol. 111 | no. 46 | 16391

SEE COMMENTARY CELL BIOLOGY

Inhibition of RUNX1 Promotes Myofibroblast Differentiation. Immunofluorescent staining of RUNX1 with or without TGF-β1 treatment in hP-MSCs showed nuclear RUNX1 expression only after TGF-β1 induction, suggesting that TGF-β1 induces nuclear accumulation of RUNX1 (Fig. 3A). To elucidate the role of RUNX1 in myofibroblast differentiation, RUNX1 expression in hP-MSCs was attenuated using two different siRNAs (Fig. 3B; SI Appendix, Figs. S12A and S14A). The prototypical genes associated with myofibroblasts, ACTA2 and TNC, exhibited a dramatic increase in RUNX1 knockdown conditions (Fig. 3D; SI Appendix, Fig. S12B). Suppression of RUNX1 also increased BMP6 and PTGS2 (also known as COX2) but not WNT11 (Fig. 3D; SI Appendix, Fig. S12B). Under coculture conditions with LNCaP expressing TGF-β1, an even greater increase in myofibroblast-associated genes was observed (Fig. 3D; SI Appendix, Fig. S12B). Engineered

N

1 ID

X1 N

Fig. 2. hP-MSCs (HPS-19I) self-organize with LNCaP cells, and paracrine TGFβ1 drives hP-MSC differentiation into myofibroblasts with a reactive stroma signature. (A) Schematic diagram of 3D organoid coculture. LNCaP cells engineered to express active TGF-β1 were cultured as organoids on inserts. hP-MSCs were cocultured with engineered LNCaP cells either in direct contact or on laminin-coated coverslips in the bottom chamber. (B) HPS-19I cells self-organize with LNCaP cells in vitro and in vivo. Note that recombined 3D organoids exhibit a core of hP-MSCs surrounded by an outer layer of LNCaP cells. (C) Immunostaining of AR and CD44/CD90 in 3D organoids indicates a periphery of LNCaP cells and a core of hP-MSCs, respectively. (D) IHC analysis of α-SMA and tenascin-C in 3D organoids consisting of hP-MSCs and engineered LNCaP cells. Note the increase in both markers in hP-MSCs when cocultured with LNCaP cells expressing TGF-β1. (E) Gene expression analysis of markers for reactive stroma and stemness, growth factors, inflammation pathway components, and transcription factors in hP-MSCs indirectly cocultured with LNCaP control or LNCaP-expressing TGF-β1. Data represent mean ± SEM. n = 3; *P < 0.05; **P < 0.01; ***P < 0.001. (Scale bars, 100 μm.)

X1

**

0

N

**

2 ***

si

4

U

**

6

si R

8

*

FB

5 4 3 2 1 0

U

Inflammatory response Transcription factors

B

PR Y1 O PO M1 U 5F 1 B M P2

TH

44 D C

Tenascin-C

***

**

U

**

**

C

**

0.5

**

R

1.0

8

Stemness

1.5

Control TGF- 1 **

IL XC 6 L1 2

TN

C TA 2 F C AP O L1 A 1 VI M C TG F FG F2

**

0.0

Tenascin-C

**

C

**

***

R

***

Growth factors

B M P TG 6 FB 1 IG F W 1 N T5 W A N T1 1

***

S2

-SMA

35 30 25 5 4 3 *** 2 1 0

***

PT G

-SMA

TGF- 1

Reactive stroma

12 10 8 6 4 2 0

C

Control

Arbitrary relative expression

D

E

A

Laminin-coated coverslip

IL

Cell culture medium (Serum-free medium containing 0.2% BSA )

si

Porous membrane (Cell culture plate insert)

Arbitrary relative expression

CD90

CD44

transcription factor-focused PCR arrays were used to identify candidate genes. Gene expression analysis of hP-MSCs from indirect cocultures showed differential regulation of transcription factor expression (SI Appendix, Dataset S2). qRT-PCR analysis was performed to validate the top six major transcription factors upregulated by TGF-β1–expressing LNCaP in hP-MSCs (SI Appendix, Fig. S10). From this analysis, we further evaluated RUNX1 and its binding partner CBFB, as well as ID1, as these exhibited higher expression levels relative to the other upregulated genes. Each of these was verified as being TGF-β1–regulated in hP-MSCs (Fig. 2E; SI Appendix, Figs. S7 and S8B). Consistent with increased mRNA, hP-MSCs treated with TGF-β1 also exhibited elevated expression of RUNX1 protein (SI Appendix, Fig. S11). RUNX1 has been identified as a key transcription factor required for definitive hematopoiesis by regulating stem/progenitor cell proliferation and differentiation (22). RUNX1 is also reported to be involved in chondrocyte differentiation (23) and was shown to be a target of TGF-β signaling (24). These findings prompted us to further study a potential functional role of RUNX1 in MSC biology and differentiation to myofibroblasts.

C

AR

In Vivo

si N

C

Epithelial cell (LNCaP cell engineered to express active TGF- 1) Stromal cell (hP-MSC)

In Vitro

Control

B

Indirect organoid coculture

TGF- 1

Direct organoid coculture

Relative expression to GAPDH

A

overexpression of RUNX1 (Fig. 3C) produced no significant increase or decrease in the effect of TGF-β1 on hP-MSCs (Fig. 3D). These results suggest that RUNX1 is permissive for myodifferentiation but may restrict complete differentiation, perhaps by maintenance of a proliferative (progenitor) status. Accordingly, we next evaluated the role of RUNX1 in mediating proliferation of MSCs. RUNX1 Is Required for MSC Cell Cycle Progression and Proliferation.

Knockdown of RUNX1 altered cell morphology. hP-MSCs with RUNX1 knockdown exhibited a more spread out, large, and flattened phenotype (Fig. 4A). Consistent with this phenotype, RUNX1 knockdown resulted in decreased CD44, POU5F1 (also known as OCT4), and BMP2 expression, yet maintained THY1 (also known as CD90) expression (Fig. 4B; SI Appendix, Fig. S14B). THY1 is a known marker of prostate cancer-associated fibroblasts (25). Of most interest, hP-MSCs with RUNX1 knockdown were restricted from proliferation (Fig. 4C). Consistent with growth arrest, RUNX1 knockdown led to decrease in expression of proliferation-associated genes (Fig. 4D; SI Appendix, Fig. S14B). Bromodeoxyuridine (BrdU) incorporation revealed that depletion of RUNX1 resulted in a decrease in the number of cells in S phase (Fig. 4E; SI Appendix, Fig. S14C). Analysis of cell cycle-related gene expression showed a significant decrease in cyclin A and B, required for both S-phase progression and G2/M-phase transition and for mitosis, respectively (26–28) (Fig. 4F; SI Appendix, Fig. S14B). No significant differences were observed in apoptotic and necrotic cell death between control and RUNX1 knockdown hP-MSCs, as analyzed by annexin V conjugate and SYTOX Red dead cell stain (SI Appendix, Fig. S15). These data suggest that attenuated RUNX1 restricts proliferation by modulating cell cycle regulators without inducing cell death. RUNX1 knockdown in marrow-derived MSCs produced nearly identical results. These cells also exhibited a more flattened phenotype and were fully growth restricted under RUNX1 knockdown

* *

A N

K

PC

I6 7 cM YC B IR C 5

0.0

**

1.0 0.5

**

*

**

B 1 N B 2 C N D C 1 C N E1

0.0

C

siNC siRUNX1

C

S 4.94%

*

0.5

2.0

N A 2

1

siRUNX1

S 18.3%

1.0

1.5

M P2

siNC

4

F

siNC siRUNX1

B

U 5F

Y1 TH

PO

44 D

2 3 Days

1.5

M

1

Arbitrary relative expression

0

C

***

**

0.5

C

BrdU :: APC

2

1.0

0.0

E

***

C

C

Stemness

1.5

4

C N

TA

2

C

0

6

C

2

D siNC siRUNX1

8

C

***

4

C

Arbitrary relative expression

**

6

Cell Number (x105)

Myofibroblast signature 8

A

siRUNX1

B

TN

siNC

Arbitrary relative expression

A

3 ** 2

*

1

N 2B K

N 2A

C D

K C D

N 1B K

C D

DNA content :: 7-AAD

C D

K

N 1A

0

Fig. 4. Suppressing RUNX1 induces hP-MSC (HPS-19I) growth and cell cycle arrest. (A) Effect of RUNX1 knockdown on hP-MSC morphology. (Scale bars, 50 μm.) (B) qRT-PCR analysis of myofibroblast and stemness genes in control and RUNX1 knockdown hP-MSCs. (C) Growth curves of hP-MSCs following transfection with control or RUNX1 siRNA. (D) qRT-PCR analysis of proliferation-related genes in control and RUNX1 knockdown hP-MSCs. (E) Representative cell cycle distribution of hP-MSCs transfected with control or RUNX1 siRNA. Cells were analyzed by flow cytometry after BrdU incorporation and 7-AAD staining at 72 h posttransfection. (F) qRT-PCR analysis of multiple cyclins and cyclin-dependent kinase inhibitors in hP-MSCs following RUNX1 knockdown. For B, D, and F, data represent mean ± SEM. For C, data represent mean ± SD; two-way ANOVA. n = 3; *P < 0.05; **P < 0.01; ***P < 0.001.

16392 | www.pnas.org/cgi/doi/10.1073/pnas.1407097111

conditions (Fig. 5A). In a similar manner, the percentage of cells in S phase was reduced (Fig. 5E). Expression of proliferation-associated genes and cell cycle control genes (Fig. 5 B and D), as well as differentiation/stemness-associated genes (Fig. 5C) exhibited a pattern consistent with those observed in hP-MSCs. To assess how RUNX1 expression affects cell proliferative or differentiation-committed states, we examined the localization of RUNX1, Ki-67, and α-SMA after RUNX1 knockdown in both hPMSCs and hBM-MSCs. Cell morphology changes and α-SMA expression were correlated with the nuclear intensity of RUNX1 staining (Fig. 6A). In control hP-MSCs, small and undifferentiated cells showed higher RUNX1 nuclear staining and lower α-SMA cytoplasmic expression (arrowheads) than large and differentiated cells (arrows). More interestingly, RUNX1-depleted hP-MSCs and hBM-MSCs revealed a myofibroblast phenotype characterized by actin polymerization, whereas controls expressed nonpolymerized actin. A similar correlation to that seen between RUNX1 and α-SMA was observed between Ki-67 and α-SMA (Fig. 6B). An increase in α-SMA in MSCs corresponded to loss of Ki-67 expression and morphology changes consistent with myofibroblast differentiation. Ki-67 negative cells in RUNX1 knockdown MSCs exhibited highly organized α-SMA (arrows). In addition, in vivo xenografts (day 3) constructed with LNCaP cells combined with RFP-expressing control or RUNX1 knockdown hP-MSCs also showed reduced Ki-67 staining in RUNX1 knockdown conditions (SI Appendix, Fig. S16). Consistent with these data, xenografts constructed with RUNX1 knockdown hP-MSCs exhibited a greater staining intensity for α-SMA (SI Appendix, Fig. S13). Overall, these results suggest that RUNX1 in MSCs is required for cell proliferation and modulates terminal differentiation to myofibroblasts (Fig. 6C). Discussion We report the first (to our knowledge) isolation and characterization of tissue-resident MSCs derived from human normal adult prostate gland (21). The prostate-derived MSCs exhibited MSCtypical cell-surface antigen profiles and were multipotent and capable of long-term culture. Previous reports of adult human tissueresident MSCs are restricted to dental (29), adenoid (30), and adipose (31) tissues. Importantly, we report here a critical role of RUNX1 in modulating MSC cell proliferation and committed differentiation to myofibroblasts. Myofibroblasts are stromal cells that are crucial for proper wound repair and the formation of reactive stroma associated with cancer, fibrosis, and vessel disease. Our data suggest that prostate-resident CD44+CD90+ MSCs are one potential source of reactive stroma myofibroblasts that coevolve with foci of prostate cancer and in benign prostatic hyperplasia. There are several reported sources of myofibroblasts and carcinoma associated fibroblasts in tumor-associated reactive stroma, including circulating MSCs, normal fibroblasts, circulating fibrocytes, and vessel-associated pericytes (1, 2, 32, 33). Our data do not exclude the involvement of these cell types. Our data further show that RUNX1 is an essential transcription factor for MSC proliferation and that RUNX1 modulates MSC differentiation to myofibroblasts. TGF-β1, overexpressed by prostate carcinoma cells, is a key inducer of myofibroblast differentiation and biology. We show here that RUNX1 expression is regulated by TGF-β1 in MSCs and that RUNX1 affects TGF-β1–stimulated gene expression during differentiation to myofibroblasts. Myofibroblasts are important in normal wound closure and contribute to the pathology of several diseases and disorders. Importantly, myofibroblasts provide a prowound repair environment that forms granulation tissue and contracts to affect wound closure. The transient appearance of myofibroblasts is beneficial in restoring and maintaining tissue homeostasis. However, excessive myofibroblast proliferation and prolonged presence is detrimental for tissue function and thereby increases the risk of fibrotic diseases and cancers in many tissues including lung, liver, kidney, and prostate (4, Kim et al.

Arbitrary relative expression C C N A 2 C C N B 1 C C N B 2 C C N D C 1 C N E1

*

**

***

E

***

1

TN C

A C TA 2

0

siNC siRUNX1

Stemness

2.0

**

1.5 1.0 ***

siNC

P2

***

U 5F 1

0.0

B M

0.5

TH Y1

5 C

2

PO

1.0 0.5

siNC siRUNX1

siNC siRUNX1

**

siRUNX1

4

**

3

***

2

**

1

BrdU :: APC

0.0

S 26.4%

S 3.73%

extracellular matrix components. Organoid coculture of MSCs and tumor cells consistently formed a self-organizing organoid with a core of MSC-derived stromal cells surrounded by tumor cells. The 3D indirect coculture model permits a more detailed understanding of the complex mechanisms underlying carcinoma cell–stromal cell crosstalk. Myofibroblasts are currently a potent target for developing antifibrotic therapies (48). Our findings provide insights to help identify potentially specific pathways. The overall key mechanisms and contributing cell types in the formation of reactive stroma have yet been fully identified. Perturbation of pathways that produce a biologically effective mass of myofibroblasts may offer potential for the development of antifibrotic and anticancer therapies via targeting the reactive stroma microenvironment.

SEE COMMENTARY

Myofibroblast signature

C D 44

1.5

IR

Days

D

0.0 B

4

7

3

*** YC

2

***

cM

1

A

0

***

0.5

I6

2

1.0

N

***

4 3

1.5

K

4

C

M

6

Arbitrary relative expression

siNC siRUNX1

PC

Cell Number (x105)

B 8

Arbitrary relative expression

A

N K D C

DNA content :: 7-AAD

Fig. 5. Depletion of RUNX1 inhibits growth and cell cycle progression in hBM-MSCs. (A) Growth curve of hBM-MSCs in control and RUNX1 knockdown conditions. (B–D) qRT-PCR analysis in hBM-MSCs transfected with control or RUNX1 siRNA at 48 h posttransfection. Gene expression profiles of proliferation-associated genes (B), myofibroblast and stemness genes (C), and cell cycle control genes (D). (E) Representative cell cycle distribution of hBM-MSCs pulse-labeled with BrdU following transfection with control or RUNX1 siRNA for 72 h. Note that loss of RUNX1 in hBM-MSCs exhibits consistent phenotype and gene expression patterns as those shown in hP-MSCs. For A, data represent mean ± SD; two-way ANOVA. For B–D, data represent mean ± SEM. n = 3; *P < 0.05; **P < 0.01; ***P < 0.001.

34, 35). Thus, reactive stroma is associated with elevated accumulation of myofibroblasts, a hallmark of fibrosis and a predictor of cancer progression. Accordingly, identifying mechanisms that affect the genesis of myofibroblasts from precursor cells may provide potential therapeutic targets in the treatment of fibrotic diseases and cancers. RUNX1 is a key transcription factor that regulates hematopoietic stem cells and hematopoiesis (36); however, its role in MSC and myofibroblast biology has not been reported. Our data show that RUNX1 expression is required for extended self-renewal and proliferation of MSCs and that RUNX1 expression is stimulated by TGF-β1. Attenuation of RUNX1 induced complete myofibroblast differentiation and quiescence of hP-MSCs, and this was potentiated by TGF-β1. Interestingly, these data show that TGF-β1 induces expression of genes associated with differentiation of myofibroblasts, yet also stimulates expression of RUNX1 that inhibits differentiation and maintains proliferation. The genesis of reactive stroma myofibroblasts depends on the proliferation of MSCs and their transit-amplifying progeny. Hence, one possible explanation for our data is that TGF-β1 induces MSC commitment to myofibroblast progenitor cells while maintaining proliferative status via elevated or sustained expression of RUNX1. This concept is consistent with previous reports that show that TGF-β1 stimulates proliferation of mesenchymal stem cells (37–39), progenitor cells, and several types of fibroblasts (40–43) and also induces myofibroblast differentiation and the fibrosis phenotype (44). TGF-β has been shown to stimulate fibroblast proliferation and subsequently induce their quiescence and differentiation to myofibroblasts in a coordinate manner (45). Additional factors, including CXCL12 (SDF-1), have also been shown to regulate the myofibroblast phenotype (46). TGF-β and CXCL12 pathways interact and regulate accumulation of myofibroblasts from resident fibroblasts in breast cancer (47). Hence, TGF-β1 acts in a coordinate and sequential manner to both stimulate proliferation of progenitors and induce their differentiation. Another advance reported here was the development of a cancer cell–MSC recombined 3D in vitro organoid coculture model. This model requires no additional growth factors, serum substitutes, or Kim et al.

Isolation of Stromal Cells from Normal Human Prostate Gland. A fresh tissue core from the peripheral zone of normal prostate was obtained from 19- and 33-y-old cadaver donors following a Baylor College of Medicine Institutional Review Board approved protocol and consents. The core was cut into discs and placed in a 96-well tissue plate containing Bfs medium [DMEM; high glucose (GIBCO) supplemented with 5% (vol/vol) FBS (HyClone), 5% (vol/vol) NuSerum (Collaborative Research), 0.5 μg/mL testosterone, 5 μg/mL insulin, 100 units/mL penicillin, and 100 μg/mL streptomycin (Sigma)]. The explants were incubated at 37 °C with 5% (vol/vol) CO2, and medium was changed every 48 h or as necessary. Stromal cells migrated out of the explant and attached to the tissue culture dish. After the cells reached confluence, the explant was moved and the cells were passaged. Cultures at passages 7–15 were used for all experiments.

siNC

A

siRUNX1

hP-MSCs

B

RUNX1 -SMA

RUNX1 -SMA

Ki-67 -SMA

Ki-67 -SMA

Ki-67 -SMA

Ki-67 -SMA

hP-MSCs

hBM-MSCs

C

TGF- 1 RUNX1 RUNX1 RUNX1

RUNX1

Cell Differentiation f RUNX1 RUNX1 Differentia proliferation

CD44 CD90+ hP-MSC hBM-MSC +

Myofibroblast -SMA, tenascin-C

Transit-amplifying cell population TGF- 1

RUNX RUNX1

Myofibroblast enriched stroma (reactive)

TGF- 1

Blocked cell proliferation

RUNX RUNX1

Differentiation

Myofibroblast depleted stroma

Fig. 6. The expression level of RUNX1 regulates cell proliferative or differentiation-committed state in hP-MSCs (HPS-19I) and hBM-MSCs. (A) Immunofluorescent staining for RUNX1 (red) and α-SMA (green) in hP-MSCs transfected with control or RUNX1 siRNA for 72 h. (Scale bars, 100 μm.) (B) Immunofluorescent staining for Ki-67 (red) and α-SMA (green) in hP-MSCs and hBM-MSCs with siRNA-mediated knockdown RUNX1 or control siRNA at 72 h posttransfection. Note that silencing RUNX1 increased polymerized α-SMA that corresponded to loss of Ki-67 expression and morphology changes consistent with myofibroblast differentiation (arrows). (Scale bars, 100 μm.) (C) Schematic diagram depicting a working hypothesis of how RUNX1 and TGF-β1 may regulate the transit amplification of MSCs to a critical mass of myofibroblasts typical of wound repair stroma and reactive stroma associated with cancer and fibrosis.

PNAS | November 18, 2014 | vol. 111 | no. 46 | 16393

CELL BIOLOGY

2A

Materials and Methods

2B

1B

N

N

K

K

D

D

C

C

C

D

K

N

1A

0

Coculture of hP-MSCs and Prostate Cancer Cells. For 3D organoid coculture, a mixture of LNCaP cells and/or either hBM-MSCs (Lonza) or hP-MSCs at different ratios was seeded onto a 12-mm Millicell-CM culture plate insert (Millipore) in a 24-well culture plate (Falcon) and incubated in LNCaP cell growth medium. After 24 h of incubation, the medium was switched to M20 medium [DMEM; high glucose (GIBCO) supplemented with 0.2% BSA, 5 μg/mL insulin, 5 μg/mL transferrin, and 5 ng/mL sodium selenite (ITS; insulin, trasnsferrin, selenium supplement, Sigma)]. After 72 h of coculture, organoids were either fixed in 4% (vol/vol) paraformaldehyde for 15 min at room temperature for paraffin embedding or immediately processed for cryo-embedding. For indirect coculture, hP-MSCs (5 × 104) were seeded on a 12-mm round coverslip coated with poly-D-lysine/ laminin (BD BioCoat) on a 24-well culture plate (Falcon) in Bfs medium. LNCaP or PC3 cells (4 × 105) were seeded onto a 12-mm Millicell-CM culture plate insert (Millipore) in a 24-well culture plate (Falcon) in growth medium. After 24 h of incubation, the insert containing LNCaP cells was transferred to the well containing hP-MSCs, and these cells were cocultured for 72 h in M20 medium. For direct coculture in monolayer, RFP-expressing hP-MSCs (1 × 106) were seeded onto a 150-mm dish (Falcon) in Bfs medium. After 24 h of incubation, an equal number of GFP-expressing engineered LNCaP cells were directly loaded onto hPMSCs and cultured for 24 h in LNCaP cell growth medium. The next day, the medium was switched to M20 medium, and cells were cocultured for 72 h. hPMSCs (RFP+) and engineered LNCaP (GFP+) cells from coculture experiments were

suspended in Ca2+/Mg2+-free Dulbecco’s PBS (GIBCO). RFP- and GFP-labeled cell sorting was performed with a BD FACSAria II cell sorter.

1. Uccelli A, Moretta L, Pistoia V (2008) Mesenchymal stem cells in health and disease. Nat Rev Immunol 8(9):726–736. 2. Barron DA, Rowley DR (2012) The reactive stroma microenvironment and prostate cancer progression. Endocr Relat Cancer 19(6):R187–R204. 3. Desmoulière A, Darby IA, Gabbiani G (2003) Normal and pathologic soft tissue remodeling: Role of the myofibroblast, with special emphasis on liver and kidney fibrosis. Lab Invest 83(12):1689–1707. 4. Gabbiani G (2003) The myofibroblast in wound healing and fibrocontractive diseases. J Pathol 200(4):500–503. 5. Rowley DR (1998-1999) What might a stromal response mean to prostate cancer progression? Cancer Metastasis Rev 17(4):411–419. 6. Tuxhorn JA, Ayala GE, Rowley DR (2001) Reactive stroma in prostate cancer progression. J Urol 166(6):2472–2483. 7. Tuxhorn JA, et al. (2002) Reactive stroma in human prostate cancer: Induction of myofibroblast phenotype and extracellular matrix remodeling. Clin Cancer Res 8(9): 2912–2923. 8. Peddareddigari VG, Wang D, Dubois RN (2010) The tumor microenvironment in colorectal carcinogenesis. Cancer Microenviron 3(1):149–166. 9. Gao Q, et al. (2011) Tumor stroma reaction-related gene signature predicts clinical outcome in human hepatocellular carcinoma. Cancer Sci 102(8):1522–1531. 10. Bremnes RM, et al. (2011) The role of tumor stroma in cancer progression and prognosis: Emphasis on carcinoma-associated fibroblasts and non-small cell lung cancer. J Thorac Oncol 6(1):209–217. 11. Dekker TJ, et al. (2013) Prognostic significance of the tumor-stroma ratio: Validation study in node-negative premenopausal breast cancer patients from the EORTC perioperative chemotherapy (POP) trial (10854). Breast Cancer Res Treat 139(2):371–379. 12. Freeman MR, Li Q, Chung LW (2013) Can stroma reaction predict cancer lethality? Clin Cancer Res 19(18):4905–4907. 13. Ayala G, et al. (2003) Reactive stroma as a predictor of biochemical-free recurrence in prostate cancer. Clin Cancer Res 9(13):4792–4801. 14. Ayala GE, et al. (2011) Determining prostate cancer-specific death through quantification of stromogenic carcinoma area in prostatectomy specimens. Am J Pathol 178(1):79–87. 15. Yanagisawa N, et al. (2007) Stromogenic prostatic carcinoma pattern (carcinomas with reactive stromal grade 3) in needle biopsies predicts biochemical recurrence-free survival in patients after radical prostatectomy. Hum Pathol 38(11):1611–1620. 16. Tuxhorn JA, McAlhany SJ, Dang TD, Ayala GE, Rowley DR (2002) Stromal cells promote angiogenesis and growth of human prostate tumors in a differential reactive stroma (DRS) xenograft model. Cancer Res 62(11):3298–3307. 17. Tuxhorn JA, McAlhany SJ, Yang F, Dang TD, Rowley DR (2002) Inhibition of transforming growth factor-beta activity decreases angiogenesis in a human prostate cancer-reactive stroma xenograft model. Cancer Res 62(21):6021–6025. 18. Yang F, Strand DW, Rowley DR (2008) Fibroblast growth factor-2 mediates transforming growth factor-beta action in prostate cancer reactive stroma. Oncogene 27(4):450–459. 19. Yang F, et al. (2005) Stromal expression of connective tissue growth factor promotes angiogenesis and prostate cancer tumorigenesis. Cancer Res 65(19):8887–8895. 20. Hinz B (2007) Formation and function of the myofibroblast during tissue repair. J Invest Dermatol 127(3):526–537. 21. Rowley DR, Tindall DJ (1987) Responses of NBT-II bladder carcinoma cells to conditioned medium from normal fetal urogenital sinus. Cancer Res 47(11):2955–2960. 22. Nimmo R, Woollard A (2008) Worming out the biology of Runx. Dev Biol 313(2): 492–500. 23. Wang Y, et al. (2005) Runx1/AML1/Cbfa2 mediates onset of mesenchymal cell differentiation toward chondrogenesis. J Bone Miner Res 20(9):1624–1636. 24. Ito Y, Miyazono K (2003) RUNX transcription factors as key targets of TGF-beta superfamily signaling. Curr Opin Genet Dev 13(1):43–47. 25. Zhao H, Peehl DM (2009) Tumor-promoting phenotype of CD90hi prostate cancerassociated fibroblasts. Prostate 69(9):991–1000.

26. Fung TK, Ma HT, Poon RY (2007) Specialized roles of the two mitotic cyclins in somatic cells: Cyclin A as an activator of M phase-promoting factor. Mol Biol Cell 18(5): 1861–1873. 27. Gong D, Ferrell JE, Jr (2010) The roles of cyclin A2, B1, and B2 in early and late mitotic events. Mol Biol Cell 21(18):3149–3161. 28. Pagano M, Pepperkok R, Verde F, Ansorge W, Draetta G (1992) Cyclin A is required at two points in the human cell cycle. EMBO J 11(3):961–971. 29. Jo YY, et al. (2007) Isolation and characterization of postnatal stem cells from human dental tissues. Tissue Eng 13(4):767–773. 30. Lee YS, et al. (2013) Isolation of mesenchymal stromal cells (MSCs) from human adenoid tissue. Cell Physiol Biochem 31(4-5):513–524. 31. Araña M, Mazo M, Aranda P, Pelacho B, Prosper F (2013) Adipose tissue-derived mesenchymal stem cells: Isolation, expansion, and characterization. Methods Mol Biol 1036:47–61. 32. Gabbiani G (1996) The cellular derivation and the life span of the myofibroblast. Pathol Res Pract 192(7):708–711. 33. Brennen WN, Chen S, Denmeade SR, Isaacs JT (2013) Quantification of Mesenchymal stem cells (MSCs) at sites of human prostate cancer. Oncotarget 4(1):106–117. 34. Hinz B, et al. (2012) Recent developments in myofibroblast biology: Paradigms for connective tissue remodeling. Am J Pathol 180(4):1340–1355. 35. Desmoulière A, Chaponnier C, Gabbiani G (2005) Tissue repair, contraction, and the myofibroblast. Wound Repair Regen 13(1):7–12. 36. Chen MJ, Yokomizo T, Zeigler BM, Dzierzak E, Speck NA (2009) Runx1 is required for the endothelial to haematopoietic cell transition but not thereafter. Nature 457(7231): 887–891. 37. Jian H, et al. (2006) Smad3-dependent nuclear translocation of beta-catenin is required for TGF-beta1-induced proliferation of bone marrow-derived adult human mesenchymal stem cells. Genes Dev 20(6):666–674. 38. Rodrigues M, Griffith LG, Wells A (2010) Growth factor regulation of proliferation and survival of multipotential stromal cells. Stem Cell Res Ther 1(4):32. 39. Walenda G, et al. (2013) TGF-beta1 does not induce senescence of multipotent mesenchymal stromal cells and has similar effects in early and late passages. PLoS ONE 8(10):e77656. 40. Chen G, Deng C, Li YP (2012) TGF-β and BMP signaling in osteoblast differentiation and bone formation. Int J Biol Sci 8(2):272–288. 41. Cheng L, Zhang C, Li D, Zou J, Wang J (2014) Transforming growth factor-β1 (TGF-β1) induces mouse precartilaginous stem cell proliferation through TGF-β receptor II (TGFRII)-Akt-β-catenin signaling. Int J Mol Sci 15(7):12665–12676. 42. Mendias CL, Gumucio JP, Lynch EB (2012) Mechanical loading and TGF-β change the expression of multiple miRNAs in tendon fibroblasts. J Appl Physiol (1985) 113(1):56–62. 43. Stouffer GA, Owens GK (1994) TGF-beta promotes proliferation of cultured SMC via both PDGF-AA-dependent and PDGF-AA-independent mechanisms. J Clin Invest 93(5): 2048–2055. 44. Desmoulière A, Geinoz A, Gabbiani F, Gabbiani G (1993) Transforming growth factorbeta 1 induces alpha-smooth muscle actin expression in granulation tissue myofibroblasts and in quiescent and growing cultured fibroblasts. J Cell Biol 122(1): 103–111. 45. Grotendorst GR, Rahmanie H, Duncan MR (2004) Combinatorial signaling pathways determine fibroblast proliferation and myofibroblast differentiation. FASEB J 18(3): 469–479. 46. Gharaee-Kermani M, et al. (2012) CXC-type chemokines promote myofibroblast phenoconversion and prostatic fibrosis. PLoS ONE 7(11):e49278. 47. Kojima Y, et al. (2010) Autocrine TGF-beta and stromal cell-derived factor-1 (SDF-1) signaling drives the evolution of tumor-promoting mammary stromal myofibroblasts. Proc Natl Acad Sci USA 107(46):20009–20014. 48. Kramann R, DiRocco DP, Humphreys BD (2013) Understanding the origin, activation and regulation of matrix-producing myofibroblasts for treatment of fibrotic disease. J Pathol 231(3):273–289.

16394 | www.pnas.org/cgi/doi/10.1073/pnas.1407097111

Differential Reactive Stroma Xenograft Model. Mice were injected subcutaneously with LNCaP cells and hP-MSCs with or without Matrigel as described previously (16). Xenografts were harvested after 3, 10, and 11 d. Statistical Analysis. Student t test or two-way ANOVA was used to determine significance between groups. For all statistical tests, P < 0.05 was considered significant. ACKNOWLEDGMENTS. We thank Drs. Owen Witte and Li Xin for the FUCRW lentiviral vector and packaging plasmids; Truong D. Dang for technical assistance; and Joel M. Sederstrom and Michael Ittmann, MD, PhD, for expert assistance. This work was supported by grants from the National Institutes of Health (NIH) (R01 CA58093 and R01 DK083293) and the Department of Defense (W81XWH-12-1-0197). This project was supported by the Cytometry and Cell Sorting Core at Baylor College of Medicine with funding from the NIH (AI036211, CA125123 and RR024574) and by the Pathology and Histology Core at Baylor College of Medicine with funding from the NIH (National Cancer Institute Grant P30CA125123).

Kim et al.

RUNX1 is essential for mesenchymal stem cell proliferation and myofibroblast differentiation.

Myofibroblasts are a key cell type in wound repair, cardiovascular disease, and fibrosis and in the tumor-promoting microenvironment. The high accumul...
1MB Sizes 0 Downloads 6 Views